The proposed research explores and characterizes expert and novice intensive care decision making within a clinical setting. The aims are to construct process tracing models of the knowledge structure and heuristics of clinicians, to develop statistical models of their judgment and to compare these with process tracing models, to evaluate framing effects and other biases on decisions, and to study the calibrations of clinicians as a function of experience. These aims are addressed in four studies whose objectives are related to the longer term goals of studying the knowledge structure and basis of human judgment in general and clinical reasoning in particular. The first set of two studies, on process tracing and statistical modeling, describes the subjects' decision strategies and weighting schemes; the second set of studies detects how these strategies and weightings are discrepant with normative theory. The proposed research has practical, methodological and theoretical implications for human judgment as well as for critical care decision making in that it addresses real-world problem solving activities using a multi-method approach that is generalizable to other decision and clinical choice environments.